The video is in Portuguese (BR), but I'll leave the content below in English, which is the same thing I covered in the video! I hope you enjoy it.
Today I will discuss two important points: SQL and NoSQL. I'll explain when you should choose one or the other, with tips to help you make the best decision for your cloud projects. Let's go!
This article was requested by @whimsicalbison
Relational Databases (SQL)
Relational databases store data in tables with rows and columns, where the data is structured and follows a rigid schema. They use SQL (Structured Query Language) to manage and query data, and each table has a well-defined relationship with other tables.
Main features
- Data Structuring is highly structured, with a predefined schema that imposes rigor on data types.
- Integrity and Normalization are focused on ensuring that data is consistent and with minimal redundancy.
- ACID Transactions guarantees Atomicity, Consistency, Isolation, and Durability, which means that transactions are safe and reliable.
- Examples: MySQL, PostgreSQL, Oracle, Microsoft SQL Server.
When to use
- Scenarios where data consistency is critical, such as financial applications, ERP systems, e-commerce, etc., where data must be accurate and consistent.
- Systems with complex relationships that relational is an excellent choice if your data model involves many relationships between entities, such as products and orders in e-commerce.
- Complex Queries when you need to perform advanced queries involving complex joins and aggregations, SQL is highly optimized for this.
Non-relational databases (NoSQL)
Non-relational databases do not follow the rigid table model of relational databases. They are more flexible, designed to handle large volumes of distributed data, and are widely used for unstructured or semi-structured data.
Types of NoSQL Models:
- Documents: When it stores data in documents (such as JSON or BSON), such as MongoDB, CouchDB.
- Columnar: When it stores data in columns rather than rows, optimizing for large volumes of data and fast queries. E.g. Cassandra, HBase.
- Key-Value: When data is stored as key and value pairs. E.g. Redis, DynamoDB.
- Graphs: Especially useful for data involving many interconnections and complex relationships, such as social networks. Such as Neo4j.
Main features:
- Schema Flexibility does not require a predefined schema, allowing easy adjustment as data changes.
- Horizontal Scalability has better scalability for large volumes of distributed data.
- High Availability has many NoSQL databases prioritize high availability, which is great for applications that require continuous operations. Examples: MongoDB, Cassandra, Redis, Neo4j.
When to use what?
Relational Database (SQL) is the best choice if:
- You have a well-defined and structured data model and need to guarantee integrity and consistency.
- You are working in a complex transaction scenario (e-commerce, banking, business management systems).
- Your queries are intensive and require complex aggregations and joins.
Non-Relational Database (NoSQL) is ideal when:
- You need to scale quickly, especially with large volumes of unstructured data.
- Your application requires high data flexibility and you need to adapt your schema dynamically.
- Availability and read/write speed are more important than strict real-time consistency (e.g. social networks, big data analysis, IoT).
Which is the best?
The choice between a relational and non-relational database depends on the specific needs of your project. There is no “best” solution in absolute terms, as both models have their advantages.
Thanks for reading and watching!
Top comments (1)
Thanks for posting this, I really appreciate it! I especially liked that you gave scenarios of when to use each and illustrating that both have their various use cases